GPT-5 mini vs Kimi K2.5
Wins 0 of 7 areas
—
Wins 4 of 7 areas
Coding · Reasoning · Images and charts · Long documents
Kimi K2.5 wins more areas, narrowly.GPT-5 mini is cheaper.
Scores updated · 34 tests both models report · How we compare
Where each one wins
Tests won in each of the seven areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- CodingWriting and fixing software04Kimi K2.54 of 4 tests
- ReasoningHard problems that need careful thinking04Kimi K2.54 of 4 tests
- Images and chartsUnderstanding pictures, charts and video02Kimi K2.52 of 2 tests
- Long documentsFinding answers in very long texts01Kimi K2.51 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 each
- FactsGetting facts right instead of making them up22Even2 each
- Following instructionsDoing exactly what it is asked11Even1 each
Long documents rests on a single test.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
GPT-5 mini costs 38% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Kimi K2.5 pulls ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+42points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+17points ahead
- Short factual questions, answered correctlySimpleQA Verified+12.7points ahead
Where GPT-5 mini pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+9.3points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+5.2points ahead
- Customer-service tasks in a simulated bankτ-Bench V3 · Banking+1.3points ahead
Every test, side by side
All 34 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingKimi K2.5
- Terminal-Bench 2.1Kimi K2.5 by 423.745.7+42
- SWE-bench VerifiedKimi K2.5 by 1759.876.8+17
- SciCodeKimi K2.5 by 103949+10
- Terminal-Bench HardKimi K2.5 by 1.533.334.8+1.5
AgentsEven
- GDPValKimi K2.5 by 3.713.517.2+3.7
- τ-Bench V3 · BankingGPT-5 mini by 1.315.514.2+1.3
ReasoningKimi K2.5
- Humanity's Last ExamKimi K2.5 by 9.221.530.7+9.2
- ARC-AGI-2Kimi K2.5 by 7.44.411.8+7.4
- GPQA DiamondKimi K2.5 by 5.182.887.9+5.1
- CritPtKimi K2.5 by 3.103.1+3.1
FactsEven
- SimpleQA VerifiedKimi K2.5 by 12.721.634.3+12.7
- AA-Omniscience · AccuracyKimi K2.5 by 10.22535.2+10.2
- AA-Omniscience · Non-hallucinationGPT-5 mini by 9.343.634.3+9.3
- Vectara HHEM hallucination ratelower is betterGPT-5 mini by 1.312.914.2+1.3
Images and chartsKimi K2.5
- LMArena · VisionKimi K2.5 by 67 rating points12021269+67 rating
- MMMU-ProKimi K2.5 by 5.370.175.4+5.3
Following instructionsEven
- IFBenchGPT-5 mini by 5.275.470.2+5.2
- Multi-ChallengeKimi K2.5 by 2.45961.4+2.4
Other results15 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- Artificial Analysis Coding IndexKimi K2.5 by 31.215.646.8+31.2
- τ²-Bench Telecom (AA run)Kimi K2.5 by 27.568.495.9+27.5
- AA Agentic IndexKimi K2.5 by 12.88.921.7+12.8
- ARC-AGI-1Kimi K2.5 by 1154.365.3+11
- swe_bench_bashKimi K2.5 by 1159.870.8+11
- AA-OmniscienceKimi K2.5 by 10-17.3-7.3+10
- vectara_answer_rateGPT-5 mini by 7.799.992.2+7.7
- HMMT 2025Kimi K2.5 by 7.687.895.4+7.6
- AA IntelligenceKimi K2.5 by 6.716.823.5+6.7
- vectara_avg_summary_lengthGPT-5 mini by 57.7 rating points169.7112+57.7 rating
- AIME 2025Kimi K2.5 by 591.196.1+5
- HMMT Nov. 2025Kimi K2.5 by 584.289.2+5
- matharena_visual_math_overallKimi K2.5 by 2.478.280.6+2.4
- frontiermath_tier_4_v1GPT-5 mini by 2.16.34.2+2.1
- vectara_factual_consistencyGPT-5 mini by 1.387.185.8+1.3
Questions people ask
Which is better, GPT-5 mini or Kimi K2.5?
Kimi K2.5 wins four of the seven areas where both have results: coding, reasoning, images and charts and long documents. GPT-5 mini wins none, but costs 38% less. They are level on agents, facts and following instructions.
Which is better for coding?
Kimi K2.5. It wins 4 of the 4 coding tests both models report; GPT-5 mini wins none.
Which is cheaper?
GPT-5 mini costs $0.25 per million input tokens and $2.00 per million output tokens; Kimi K2.5 costs $0.60 and $3.00. That makes GPT-5 mini about 38% cheaper for the same work.
How do you compare the two?
We use the 34 benchmark tests both models have published scores on. The verdict counts the 19 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 15 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.